unable to start embedded Tomcat

1,059 views
Skip to first unread message

Sonny Heer

unread,
Aug 10, 2017, 10:25:14 PM8/10/17
to Kylo Community
Hi users,

While creating a transformation feed - during step 4.  i get the following error in spark shell:


2017-08-11 02:16:04 INFO  launcher-proc-2:SparkShellApp:61 -  diagnostics: User class threw exception: org.springframework.context.ApplicationContextException: Unable to start embedded container; nested exception is org.springframework.boot.context.embedded.EmbeddedServletContainerException: Unable to start embedded Tomcat




Using 0.8.2 on HDP 2.4


Thanks

Sonny Heer

unread,
Aug 10, 2017, 10:39:44 PM8/10/17
to Kylo Community
Here is what is in the yarn logs:

java.lang.ClassNotFoundException: org.datanucleus.api.jdo.JDOPersistenceManagerFactory


classpath issue.  where does that class need to be loaded from?

Sonny Heer

unread,
Aug 11, 2017, 1:22:53 AM8/11/17
to Kylo Community
In case someone else is running into this.  Had to add --jars in spark.properties.  There might be an issue elsewhere still though...

spark.shell.sparkArgs=--jars /usr/hdp/current/hive-server2/lib/datanucleus-api-jdo-3.2.6.jar,/usr/hdp/current/hive-server2/lib/datanucleus-core-3.2.10.jar,/usr/hdp/current/hive-server2/lib/datanucleus-rdbms-3.2.9.jar



Now the spark job cannot find db/table.  although when I go to "tables" I'm able to view tables and preview.  My guess is its a different type of connection there vs spark job w/kerberos.  Any ideas?

Greg Hart

unread,
Aug 11, 2017, 12:50:40 PM8/11/17
to Kylo Community
Hi Sonny,

Could you try setting this property in spark.properties?
spark.shell.files=/etc/hive/conf/hive-site.xml


Sonny Heer

unread,
Aug 11, 2017, 12:55:21 PM8/11/17
to Kylo Community
Hi Greg,

I was just about to respond.  I added: 

spark.shell.sparkArgs=--files /usr/hdp/current/spark-client/conf/hive-site.xml


Got past that issue with adding the hive site file.



I'm not sure how to debug when my transformation continues to run.  e.g. A simple transform on a single Hive table without doing any spark functions.  Timer is set to run every 20 minutes.  It starts the feed, and it just keeps running.  I would expect this feed to take less than a couple minutes.


There is a yarn application started and running with my user name (AD).  Nothing else is wrong in the logs...just random updates.  

any ideas?

Sonny Heer

unread,
Aug 11, 2017, 4:03:19 PM8/11/17
to Kylo Community
I've opened another thread for the issue with transformation jobs not finishing.
Reply all
Reply to author
Forward
0 new messages